WO2019205978A1

WO2019205978A1 - Library robot positioning and navigation method

Info

Publication number: WO2019205978A1
Application number: PCT/CN2019/082719
Authority: WO
Inventors: 刘佳; 陈力军; 黄嘉琪; 张晓聪; 李珍珠; 严颖丽; 郁裕杰; 陈星宇
Original assignee: 南京大学
Priority date: 2018-04-28
Filing date: 2019-04-15
Publication date: 2019-10-31
Also published as: CN108762251A; CN108762251B

Abstract

Disclosed is a library robot positioning and navigation method, comprising: step 1, building a fuzzy inference system to obtain an inference rule; step 2, saving ID values of RFID tags which a library robot passes by in a predetermined path into a list by order; step 3, determining an ID value of an RFID tag at which the library robot needs to arrive in a next step; step 4, using two RFID reading devices mounted on the library robot to separately read RFID tags placed at the edge of a shelf, and using a quintuple Tag=<no, id, φ, r, t> to record same; step 5, performing data pre-processing on information in the quintuple Tag; and step 6, inputting the pre-processed data into the inference rule to obtain the current position information of the library robot, determining whether the library robot already arrives at a destination, and confirming a heading direction in a next step for implementing positioning and navigation functions.

Description

一种图书馆机器人定位导航方法Library robot positioning navigation method

技术领域Technical field

本发明涉及超高频RFID技术领域，尤其涉及一种图书馆机器人定位导航方法。The invention relates to the field of ultra high frequency RFID technology, in particular to a library robot positioning and navigation method.

背景技术Background technique

RFID(Radio Frequency Identification，射频识别)是一种非接触式的自动识别技术，它通过射频电磁信号自动识别目标对象并获取存储的相关数据，不需要额外人工操作。与传统的一维码(条形码)和二维码相比，RFID具有无需识别***与扫描目标间建立机械或光学接触的优点。目前，RFID技术正逐步地应用于仓储管理、物流服务以及身份识别等行业，极大的提高了管理的效率，是未来构建智能化基础设施的重要组成部分。RFID (Radio Frequency Identification) is a non-contact automatic identification technology that automatically recognizes a target object and acquires stored related data through radio frequency electromagnetic signals without additional manual operations. Compared to conventional one-dimensional codes (barcodes) and two-dimensional codes, RFID has the advantage of not requiring mechanical or optical contact between the recognition system and the scanning target. At present, RFID technology is gradually being applied to industries such as warehousing management, logistics services, and identification, which greatly improves the efficiency of management and is an important part of building an intelligent infrastructure in the future.

当今大多数图书馆的管理***主要采用“安全磁条”+“条形码”的技术手段。这种方式成本较低，实施方便，但在长时间实践过程中发现存在缺陷。随着RFID技术的不断进步以及***格的降低，其逐渐在图书馆领域进行使用，特别是超高频RFID技术，由于其可读范围广、穿透性强等优点，为图书馆智能管理提供了更好的解决方案。将超高频RFID技术与图书馆机器人技术相结合，可以极大的减少许多需要人工进行的繁重工作，例如图书盘点、图书上架等。以图书盘点功能为例，当图书馆机器人在执行任务时，需要不断在书架间移动扫描。由于机器人的运行功能需求和空间限制，要求图书馆机器人具有快速准确进行定位导航的能力。而目前，在移动图书馆机器人定位导航领域，主要利用能感知场景环境信息特征的传感器，如：激光雷达、红外线、摄像头等，通过在图书馆机器人移动过程中不断采集场景信息，构建地图，综合利用里程计等辅助工具进行定位导航。而上述的定位导航算法往往存在诸多问题。首先，这些定位导航算法过程复杂，需要在高维空间下进行大量运算，因此其定位的频率受到限制，使得机器人无法及时获知自身的位姿信息；其次，现有的定位导航算法往往基于特征匹配或直接公式计算的方式进行定位，当机器人的工作场景高度同构化，例如：图书馆中书架的外观高度相似，或是行进路线复杂、工作空间比较狭窄的情况下，现有的定位导航算法常常出现由于特征匹配失败从而导致定位出现跳变，甚至导致出现定位失效的情况。因此，现有的定位导航算法不适用于在同构化程度高、地形复杂狭窄的图书馆环境下应用。最后，使用这些算法必须为图书馆机器人增设新的传感器，极大的提高了成本。Most of today's library management systems mainly use the "safe magnetic stripe" + "barcode" technology. This method is low in cost and easy to implement, but it has been found to be defective during a long period of practice. With the continuous advancement of RFID technology and the reduction of cost and price, it is gradually used in the field of libraries, especially UHF RFID technology. Due to its wide readable range and strong penetrability, it provides library intelligent management. A better solution. Combining UHF RFID technology with library robotics can greatly reduce the amount of heavy work that needs to be done manually, such as book inventory and book shelves. Taking the book inventory function as an example, when the library robot is performing tasks, it is necessary to continuously scan and scan between the bookshelves. Due to the operational requirements and space constraints of robots, library robots are required to have the ability to quickly and accurately position and navigate. At present, in the field of mobile library robot positioning and navigation, sensors that can sense the characteristics of the scene environment, such as: laser radar, infrared, camera, etc., continuously collect scene information during the movement of the library robot, construct a map, and integrate Use navigation tools such as odometers for navigation. The above-mentioned positioning and navigation algorithms often have many problems. First of all, these positioning and navigation algorithms are complex and require a large number of operations in high-dimensional space. Therefore, the frequency of positioning is limited, so that the robot can not know its own pose information in time. Secondly, the existing positioning and navigation algorithms are often based on feature matching. Or direct formula calculation for positioning, when the robot's working scene is highly isomorphic, for example, the height of the bookshelf in the library is highly similar, or the travel route is complex and the working space is relatively narrow, the existing positioning navigation algorithm It is often the case that the positioning fails due to the failure of the feature matching, and even the positioning failure occurs. Therefore, the existing positioning and navigation algorithms are not suitable for application in a library environment with high degree of homology and complex terrain. Finally, the use of these algorithms must add new sensors to the library robot, greatly increasing the cost.

发明内容Summary of the invention

发明目的：本发明所要解决的技术问题是针对现有技术的不足，提供基于超高频RFID技术的定位导航方法，利用模糊逻辑***和RFID技术中的RSSI(Received Signal Strength Indication接收的信号强度指示)值和相位差值，进行图书馆机器人自动定位导航。OBJECT OF THE INVENTION The technical problem to be solved by the present invention is to provide a positioning and navigation method based on UHF RFID technology, and to utilize RSSI (Received Signal Strength Indication received signal strength indication) in a fuzzy logic system and an RFID technology. ) The value and phase difference are used for automatic positioning and navigation of the library robot.

为了解决上述技术问题，本发明公开了一种图书馆机器人定位导航方法，具体步骤如下：In order to solve the above technical problem, the present invention discloses a library robot positioning and navigation method, and the specific steps are as follows:

步骤1，建立模糊推理***，得到对应的推理规则；Step 1: Establish a fuzzy inference system to obtain a corresponding inference rule;

步骤2，将图书馆机器人预定路径中经过的RFID标签的ID信息按顺序存入列表中； Step 2, the ID information of the RFID tag passing through the predetermined path of the library robot is sequentially stored in the list;

步骤3，确定图书馆机器人下一步需到达的目标RFID标签的ID信息，并计算出图书馆机器人的前进方向，图书馆机器人开始朝下一步需到达的目标RFID标签所在位置移动； Step 3, determining the ID information of the target RFID tag that the library robot needs to reach next, and calculating the forward direction of the library robot, and the library robot starts moving toward the position of the target RFID tag to be reached next;

步骤4，图书馆机器人移动过程中，使用装置于图书馆机器人上的两个RFID读取设备分别读取放置于书架边沿的RFID标签，第一个RFID读取设备的天线记为天线1，第二个RFID读取设备的天线记为天线2，RFID读取设备读取的信号包括：RFID读取设备的标号信息no，RFID标签的ID信息id，两个RFID读取设备分别采集到当前读取RFID标签的相位差

及RFID标签的RSSI值r，以及采集到该RFID标签的时间戳t，用五元组

记录；no取值为1或2，为1时表示第一个RFID读取设备，为2时表示第二个RFID读取设备； Step 4: During the movement of the library robot, the two RFID reading devices on the library robot are respectively used to read the RFID tags placed on the edge of the bookshelf, and the antenna of the first RFID reading device is recorded as the antenna 1, the first The antennas of the two RFID reading devices are recorded as antennas 2. The signals read by the RFID reading device include: label information no of the RFID reading device, ID information id of the RFID tag, and two RFID reading devices respectively collect the current reading. Take the phase difference of the RFID tag

And the RSSI value r of the RFID tag, and the time stamp t of the RFID tag, using the quintuple

Record; no value is 1 or 2, when it is 1, it means the first RFID reading device, when it is 2, it means the second RFID reading device;

步骤5，对于五元组Tag中的信息进行数据预处理；Step 5: Perform data preprocessing on information in the quintuple tag;

步骤6，将预处理后的五元组Tag中的信息输入步骤1得到的推理规则中，得到关于图书馆机器人的当前位置信息，并根据步骤2中的列表判断是否已经到达目标RFID标签所在位置，从而确认下一步的前进方向，以实现定位导航功能。 Step 6, the information in the preprocessed quintuple Tag is input into the inference rule obtained in step 1, and the current location information about the library robot is obtained, and according to the list in step 2, it is determined whether the location of the target RFID tag has been reached. In order to confirm the direction of the next step, to achieve the positioning navigation function.

其中，步骤1包括：Among them, step 1 includes:

步骤1-1，根据模糊逻辑方法，基于RFID信号的传播特性，建立两组单输入-单输出的曼丹尼模糊推理***。模糊逻辑是一种模仿人脑的不确定性概念判断和推理思维方式。应用模糊推理***，可以对模型未知或不能确定的描述***进行有效描述和计算。由于RFID信号在传输过程中容易受周围环境影响，因而无法通过精确的公式描述其物理特性，如RSSI值等。因此，可以为其建立其对应的模糊推理***。而后通过大量的实验数据对***进行修正，使得它对RFID数据的描述更加精确。在本方法中，需要建立双输入-单输出的模糊推理***，输入变量为同一RFID读取设备对同一RFID标签读取的RFID标签RSSI值和相位差

输出变量为图书馆机器人相对于RFID标签的方位角α；并得出***对应的隶属函数。通过隶属函数，可以建立起

的关系； Step 1-1, based on the fuzzy logic method, based on the propagation characteristics of the RFID signal, establish two sets of single input-single output Mandani fuzzy inference system. Fuzzy logic is a way of imitating the uncertainty and conceptual thinking of the human brain. Using the fuzzy inference system, the description system with unknown or undetermined model can be effectively described and calculated. Since the RFID signal is easily affected by the surrounding environment during transmission, it is impossible to describe its physical characteristics such as the RSSI value by an accurate formula. Therefore, it can establish its corresponding fuzzy inference system. The system is then modified with a large amount of experimental data to make it more accurate in describing the RFID data. In this method, a dual input-single output fuzzy inference system is needed, and the input variable is the RSSI value and phase difference of the RFID tag read by the same RFID reading device for the same RFID tag.

The output variable is the azimuth angle α of the library robot relative to the RFID tag; and the membership function corresponding to the system is obtained. Through the membership function, it can be established

Relationship;

步骤1-2，根据步骤1-1中建立的模糊推理***，得出如下对应的“IF-THEN”模糊逻辑推理规则：Step 1-2: According to the fuzzy inference system established in step 1-1, the following corresponding "IF-THEN" fuzzy logic inference rules are obtained:

If RSSI is High and

is Low Thenαis Low，表示如果RSSI值是高，并且相位差

值是低，则方位角α值是低； If RSSI is High and

Is Low Thenαis Low, indicating if the RSSI value is high and the phase difference is

If the value is low, the azimuth alpha value is low;

If RSSI is High and

is High Thenαis Medium，表示如果RSSI值是高，并且相位差

值是高，则方位角α值是中等； If RSSI is High and

Is High Thenαis Medium, indicating if the RSSI value is high and the phase difference is

If the value is high, the azimuth alpha value is medium;

If RSSI is Low and

is High Thenαis Medium，表示如果RSSI值是低，并且相位差

值是高，则方位角α值是中等； If RSSI is Low and

Is High Thenαis Medium, indicating if the RSSI value is low and the phase difference is

If the value is high, the azimuth alpha value is medium;

If RSSI is Low and

is Low Thenαis High，表示如果RSSI值是低，并且相位差

值是低，则方位角α值是高； If RSSI is Low and

Is Low Thenαis High, indicating if the RSSI value is low and the phase difference is

If the value is low, the azimuth alpha value is high;

RSSI值、相位差

值属于High还是Low，根据它们隶属于的函数进行判断(每个具体的环境和设备都有不同的隶属函数，参数和形式都是不同的，得根据具体的情况进行调整)； RSSI value, phase difference

Whether the value belongs to High or Low, according to the function they belong to (the specific environment and equipment have different membership functions, the parameters and forms are different, and can be adjusted according to the specific situation);

例如，假设RSSI的值为-45，根据RSSI值的隶属度函数计算，若RSSI值对于“高”的隶属度大于对于“低”的隶属度，那么RSSI值应隶属于“高”；反之，RSSI值隶属于“低”。同理，对于相位差

进行同样的操作。当得到了RSSI值以及相位差

的具体属性后，就可以根据推理规则判断具体适用于哪一条规则。假设满足规则4(If RSSI is Low and

is Low Thenαis High)，而后，根据方位角α的隶属度函数，确定方位角α在属性为“高”时的模糊集合，利用加权平均法，就可以计算出α的具体值。 For example, if the value of RSSI is -45, based on the membership function of the RSSI value, if the membership of the RSSI value for "high" is greater than the membership for "low", then the RSSI value should be subordinate to "high"; RSSI values are subject to "low". Similarly, for phase difference

Do the same. When the RSSI value and the phase difference are obtained

After the specific attributes, you can determine which rule to apply to according to the inference rules. Assume that Rule 4 is satisfied (If RSSI is Low and

Is Low Thenαis High), and then, according to the membership function of the azimuth angle α, the fuzzy set when the azimuth α is “high” is determined, and the specific value of α can be calculated by the weighted average method.

步骤2包括： Step 2 includes:

步骤2-1，在书架边沿每隔一定距离粘贴RFID标签，RFID标签的ID信息包括书架号、层数以及这一层的标号，按一定规则对所有RFID标签进行编号。例如，第10书架第1层第5个标签，其对应编号为“100105”。而后，将RFID标签的ID信息与标签的具***置坐标信息以及标签的相邻关系(即标签左右相邻的两个标签的ID信息)进行绑定，将绑定结果存入数据库中；本发明中的坐标系就是以实际的图书馆场景进行实地测量，先确定书架的坐标，再确定每个标签的坐标。In step 2-1, the RFID tag is pasted at a certain distance on the edge of the bookshelf. The ID information of the RFID tag includes the bookshelf number, the number of layers, and the label of the layer, and all the RFID tags are numbered according to a certain rule. For example, the fifth label on the first floor of the 10th bookshelf, the corresponding number is "100105". Then, the ID information of the RFID tag is bound to the specific location coordinate information of the tag and the adjacent relationship of the tag (ie, the ID information of the two tags adjacent to the left and right sides of the tag), and the binding result is stored in the database; The coordinate system in the field is to take the actual library scene for field measurement, first determine the coordinates of the bookshelf, and then determine the coordinates of each label.

步骤2-2，通过查询数据库，确定距离图书馆机器人预定路径的起始位置最近的RFID标签和终止位置最近的RFID标签，并分别记为Tag _start和Tag _end； Step 2-2, by querying the database, determining the RFID tag closest to the starting position of the predetermined path of the library robot and the RFID tag closest to the end position, and respectively recording them as Tag _start and Tag _end ;

步骤2-3，通过DFS(Depth First Search)深度优先算法，寻找以Tag _startRFID标签为起点，以Tag _endRFID标签为终点的最短路径，并将路径上经过的所有RFID标签的ID信息按照起点到终点的顺序储存入列表中。 Step 2-3: Through the DFS (Depth First Search) depth-first algorithm, find the shortest path starting from the Tag _start RFID tag and ending with the Tag _end RFID tag, and follow the ID information of all the RFID tags passing through the path. The order to the end point is stored in the list.

步骤3包括：将步骤2所得的列表中第一个RFID标签作为图书馆机器人下一步需到达的目标RFID标签。 Step 3 includes: using the first RFID tag in the list obtained in step 2 as the target RFID tag that the library robot needs to reach next.

步骤5包括： Step 5 includes:

步骤5-1，根据RFID标签的ID信息id的不同，将id为步骤3中所得的目标RFID标签id的Tag数据划分成新的子集，并按照时间戳t的值升序排序。由于两个RFID读取天线受同一RFID阅读器控制，因而，两个RFID读取天线在一次读取过程中的时间戳t是相同的；Step 5-1: According to the ID information id of the RFID tag, the tag data of the target RFID tag id obtained in step 3 is divided into new subsets, and sorted in ascending order according to the value of the time stamp t. Since the two RFID reading antennas are controlled by the same RFID reader, the time stamps t of the two RFID reading antennas during one reading are the same;

步骤5-2，将新得到的子集按照不同的RFID读取设备标号no分成两个新的子集，每个子集内将步骤4所得的所有五元组数据Tag按时间戳t的值升序排序，将排序结果分别保存在四元组

和

中，Data1中no取值为1，Data2中no取值为2，r ₁表示通过第一个RFID读取设备采集到的当前读取RFID标签的RSSI值信息，

表示通过第一个RFID读取设备采集到的当前读取RFID标签的相位差，r ₂表示通过第二个RFID读取设备采集到的当前读取RFID标签的RSSI值信息，

表示通过第二个RFID读取设备采集到的当前读取RFID标签的相位差。 Step 5-2, the newly obtained subset is divided into two new subsets according to different RFID reading device labels no, and all the quintuple data tags obtained in step 4 are sorted by the value of the time stamp t in each subset. Sort, save the sort results in the quad

with

The value of no in Data1 is 1, and the value of no in Data2 is 2, and r ₁ represents the RSSI value information of the currently read RFID tag collected by the first RFID reading device.

Represents a RFID tag reader has a phase difference by the first device to collect a RFID, r ₂ represents the RSSI value of the current information of the RFID tag read equipment collected through the second RFID read,

Indicates the phase difference of the currently read RFID tag collected by the second RFID reading device.

步骤6包括： Step 6 includes:

步骤6-1，将步骤5中所得的两个四元组Data1和Data2，依据步骤1中所得的推理规则和隶属函数，分别计算得到图书馆机器人的当前位置信息Pos ₁＝<id ₃、α ₁>，Pos ₂＝<id ₄、α ₂>，Pos ₁表示根据第一个四元组Data1计算得到的图书馆机器人的当前位置信息，其中，id ₃表示根据第一个四元组Data1计算得到的当前距离机器人最近的RFID标签的ID信息，α ₁表示根据第一个四元组Data1计算得到的当前机器人相对于该RFID标签的方位角；Pos ₂表示根据第二个四元组Data2计算得到的图书馆机器人的当前位置信息，其中，id ₄表示根据第二个四元组Data2算得到的当前距离机器人最近的RFID标签的ID信息，α ₂表示根据第二个四元组Data2计算得到的当前机器人相对于该RFID标签的方位角； Step 6-1, the two quaternions Data1 and Data2 obtained in step 5 are respectively calculated according to the inference rules and membership functions obtained in step 1, and the current position information of the library robot is calculated Pos ₁ = <id ₃ , α ₁ >, Pos ₂ =<id ₄ , α ₂ >, Pos ₁ represents the current position information of the library robot calculated according to the first quad PDF1, wherein id ₃ represents the calculation according to the first quad Data1 Obtaining the ID information of the RFID tag closest to the robot from the current distance, α ₁ represents the azimuth of the current robot relative to the RFID tag calculated according to the first quad DS1; Pos ₂ represents the calculation according to the second quad Data2 The current location information of the obtained library robot, wherein id ₄ represents the ID information of the RFID tag closest to the current distance robot calculated according to the second quaternion Data2, and α ₂ represents the calculation according to the second quaternion Data2 The azimuth of the current robot relative to the RFID tag;

步骤6-2，根据步骤6-1中所得位置信息Pos ₁和Pos ₂，依据如下公式更新图书馆机器人的方位角θ _rotation： Step 6-2, according to the position information Pos ₁ and Pos ₂ obtained in step 6-1, the azimuth angle θ _{rotation of the} library robot is updated according to the following formula:

θ _rotation＝θ _{previous-＞cuurent}+θ _{current-＞goal}.........................(1) θ _rotation =θ _{previous->cuurent} +θ _{current->goal} .........................(1)

其中，(x _previous,y _previous)为上一次定位时获得的图书馆机器人位置信息，x _previous,y _previous分别表示上一次定位时获得的图书馆机器人位置横坐标和纵坐标，(x _goal,y _goal)为图书馆机器人的目标位置，x _goal,y _goal分别表示目标位置横坐标和纵坐标，由目标RFID标签的位置确定，(x _current,y _current)为图书馆机器人当前位置，x _current,y _currentl分别表示当前位置横坐标和纵坐标，d为图书馆机器人与书架边沿间的垂直距离，δ为噪声误差，θ _{previous-＞current}是指当前位置与上一次定位位置之间的方向角，θ _{current-＞goal}是指目标位置与当前位置之间的方向角，x _id1、y _id1分别表示第一个四元组Data1计算得到的图书馆机器人的当前位置横坐标和纵坐标，x _id2、y _id2分别表示第二个四元组Data2计算得到的图书馆机器人的当前位置横坐标和纵坐标； Where (x _previous , y _previous ) is the location information of the library robot obtained during the _previous positioning, x _previous , y _previous respectively represent the horizontal and vertical coordinates of the library robot position obtained during the _previous positioning, (x _goal , y _goal) libraries for the target position of the robot, x _{_goal,} y _goal target position respectively the abscissa and ordinate, determined by the position of the target RFID tag, (x _{_current,} y _current) to the current position of the robot library, x _current, y _currentl respectively represents the horizontal and vertical coordinates of the current position, d is the vertical distance between the library robot and the edge of the bookshelf, δ is the noise error, and θ _{previous->current} refers to the direction angle between the current position and the previous positioning position. θ _{current->goal} is the direction angle between the target position and the current position, and x _id1 and y _id1 respectively represent the current position of the library robot calculated by the first quad of Data1, the abscissa and the ordinate, x _id2 , y _id2 respectively represents the current position abscissa and ordinate of the library robot calculated by the second quaternary Data2;

步骤6-3，根据步骤6-2中所得位置信息(x _current,y _current)判断图书馆机器人是否已到达步骤3所得下一步需到达的目标RFID标签所在位置，若满足(x _current,y _current)的横坐标和纵坐标的对应差值的绝对值与(x _goal,y _goal)的横坐标和纵坐标的对应差值的绝对值都小于threshold，则判断已到达目标位置。若已到达，则将步骤2中列表标记的下一步需到达的目标目标RFID标签删除，继续进行步骤6-4，否则，返回执行步骤4； Step 6-3, according to the position information (x _current , y _current ) obtained in step 6-2, it is determined whether the library robot has reached the position of the target RFID tag to be reached in the next step, if it satisfies (x _current , y _current If the absolute value of the corresponding difference between the abscissa and the ordinate of the abscissa and the ordinate is less than the absolute value of the corresponding difference between the abscissa and the ordinate of (x _goal , y _goal ), it is judged that the target position has been reached. If it has arrived, delete the target target RFID tag to be reached in the next step of the list mark in step 2, and proceed to step 6-4; otherwise, return to step 4;

步骤6-4，判断步骤2所得列表中是否还存在RFID标签ID，若存在，则返回执行步骤3，否则，则判断图书馆机器人已到达预定最终位置。In step 6-4, it is determined whether the RFID tag ID still exists in the list obtained in step 2. If yes, return to step 3, otherwise, it is determined that the library robot has reached the predetermined final position.

本发明首次将模糊逻辑引入到移动图书馆机器人领域，并高效准确的实现了定位导航功能，从而使得图书馆机器人可以更好的实现上层功能，如：图书盘点等，并且有效的利用了图书馆的特殊环境，减少了成本。The invention introduces fuzzy logic into the field of mobile library robot for the first time, and realizes the positioning and navigation function efficiently and accurately, so that the library robot can better realize the upper layer functions, such as: book inventory, etc., and effectively utilizes the library. The special environment reduces costs.

有益效果：通过本方法可以避免为图书馆机器人增加昂贵的传感器，同时避免了低效复杂的传统定位导航算法，解决了图书馆机器人在移动过程中出现的定位失效、耗费计算资源等问题；同时，为上层功能的效果改善提供了重要的基础。Beneficial effects: By using this method, it is possible to avoid adding expensive sensors to the library robot, and avoiding the inefficient and complicated traditional positioning and navigation algorithm, and solving the problems of positioning failure and computational resources incurred by the library robot in the moving process; , provides an important basis for the improvement of the upper layer function.

附图说明DRAWINGS

下面结合附图和具体实施方式对本发明做更进一步的具体说明，本发明的上述或其它方面的优点将会变得更加清楚。The advantages of the above or other aspects of the present invention will become more apparent from the detailed description of the invention.

图1是图书馆机器人与书架的位置俯视示意图。Figure 1 is a top plan view of the location of the library robot and the bookshelf.

图2是RSSI值与相位α的关系图。2 is a graph showing the relationship between the RSSI value and the phase α.

图3是相位差

与相位α的关系图。 Figure 3 is the phase difference

Diagram of the relationship with phase α.

图4a是模糊逻辑***RSSI隶属函数。Figure 4a is a fuzzy logic system RSSI membership function.

图4b是输入相位差的隶属函数。Figure 4b is a membership function of the input phase difference.

图4c是输出相位α的隶属函数。Figure 4c is a membership function of the output phase a.

图5是本发明的流程图。Figure 5 is a flow chart of the present invention.

具体实施方式detailed description

下面结合附图及实施例对本发明做进一步说明。The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

如图5所示，本发明公开了一种图书馆机器人定位导航方法，具体步骤如下：As shown in FIG. 5, the present invention discloses a library robot positioning and navigation method, and the specific steps are as follows:

步骤4，如图1所示是图书馆机器人与书架的位置俯视示意图，图1中，1是书架边沿，2是RFID标签，3是RFID天线，4是机器人底盘，图书馆机器人移动过程中，使用装置于图书馆机器人上的两个RFID读取设备分别读取放置于书架边沿的RFID标签，RFID读取设备读取的信号包括：RFID读取设备的标号信息no，RFID标签的ID信息id，两个RFID读取设备分别采集到当前读取RFID标签的相位差

记录；no取值为1或2，为1时表示第一个RFID读取设备，为2时表示第二个RFID读取设备； Step 4, as shown in Figure 1 is a schematic view of the position of the library robot and the bookshelf. In Figure 1, 1 is the edge of the bookshelf, 2 is the RFID tag, 3 is the RFID antenna, 4 is the robot chassis, and the library robot moves. The two RFID reading devices on the library robot are respectively used to read the RFID tags placed on the edge of the bookshelf. The signals read by the RFID reading device include: the label information no of the RFID reading device, and the ID information id of the RFID tag. The two RFID reading devices respectively collect the phase difference of the currently read RFID tag

其中，步骤1包括：Among them, step 1 includes:

Relationship;

If RSSI is High and

is Low Thenαis Low If RSSI is High and

Is Low Thenαis Low

If RSSI is High and

is High Thenαis Medium If RSSI is High and

Is High Thenαis Medium

If RSSI is Low and

is High Thenαis Medium If RSSI is Low and

Is High Thenαis Medium

If RSSI is Low and

is Low Thenαis High。 If RSSI is Low and

Is Low Thenαis High.

步骤2包括： Step 2 includes:

步骤2-1，在书架边沿每隔一定距离粘贴RFID标签，RFID标签的ID信息包括书架号、层数以及这一层的标号，按一定规则对所有RFID标签进行编号。例如，第10书架第1层第5个标签，其对应编号为“100105”。而后，将RFID标签与书架位置坐标绑定，将绑定结果存入数据库中；In step 2-1, the RFID tag is pasted at a certain distance on the edge of the bookshelf. The ID information of the RFID tag includes the bookshelf number, the number of layers, and the label of the layer, and all the RFID tags are numbered according to a certain rule. For example, the fifth label on the first floor of the 10th bookshelf, the corresponding number is "100105". Then, the RFID tag is bound to the coordinate position of the bookshelf, and the binding result is stored in the database;

步骤2-2，通过查询数据库，确定距离图书馆机器人预定路径的起始位置和终止位置最近的两个RFID标签，并分别记为Tag _start和Tag _end； Step 2-2, by querying the database, determining two RFID tags that are closest to the start position and the end position of the predetermined path of the library robot, and respectively record them as Tag _start and Tag _end ;

步骤2-3，通过DFS(Depth First Search)深度优先算法，寻找以Tag _start为起点，以Tag _end为终点的最短路径。将路径上经过的所有RFID标签的ID信息按找起点到终点的顺序储存入列表中。 Step 2-3: Through the DFS (Depth First Search) depth-first algorithm, find the shortest path starting from the Tag _start and ending with the Tag _end . The ID information of all the RFID tags passing through the path is stored in the list in the order of finding the start point to the end point.

步骤5包括： Step 5 includes:

和

with

Representing the phase difference of the currently read RFID tag collected by the first RFID reading device, and r ₂ representing the RSSI value information of the currently read RFID tag collected by the second RFID reading device,

步骤6包括： Step 6 includes:

θ _rotation＝θ _{previous-＞cuurent}+θ _{current-＞goal}..................(1) θ _rotation =θ _{previous->cuurent} +θ _{current->goal} ..................(1)

步骤6-3，根据步骤6-2中所得位置信息(x _current,y _current)判断图书馆机器人是否已到达步骤3所得下一步需到达的目标RFID标签所在位置，若满足(x _current,y _current)的横坐标和纵坐标的对应差值的绝对值与(x _goal,y _goal)的横坐标和纵坐标的对应差值的绝对值都小于threshold，则判断已到达目标位置。若已到达，则将步骤2中列表标记的下一步需到达的目标目标RFID标签删除，继续进行步骤6-4，否则，返回执行步骤4； Step 6-3, according to the position information (x _current , y _current ) obtained in step 6-2, it is determined whether the library robot has reached the position of the target RFID tag to be reached in the next step, if it satisfies (x _current , y _current ) corresponding to the absolute value of the difference between the abscissa and ordinate abscissa and the absolute value of (x _{_goal,} y _goal) and the corresponding ordinate differences are less than threshold, it is determined that the target position has been reached. If it has arrived, delete the target target RFID tag to be reached in the next step of the list mark in step 2, and proceed to step 6-4; otherwise, return to step 4;

实施例1Example 1

RFID技术主要由三部分构成：天线，阅读器和RFID标签。RFID标签从启动能源上区分共有三类，其中最重要的一种是无源RFID标签，RFID标签本身不具备能源，只通过反射从天线发射的射频信号来传播数据。反射的信号可以通过多种物理属性来描述，RSSI值和相位角就是其中之二。根据公式，距离与RSSI值的关系如下：RFID technology consists of three main components: antennas, readers and RFID tags. There are three types of RFID tags from the starting energy source. The most important one is the passive RFID tag. The RFID tag itself does not have energy and only transmits the data by reflecting the RF signal transmitted from the antenna. The reflected signal can be described by a variety of physical properties, and the RSSI value and phase angle are two of them. According to the formula, the relationship between distance and RSSI value is as follows:

d＝10^((abs(RSSI)-A)/(10*n)).....(6)d=10^((abs(RSSI)-A)/(10*n)).....(6)

其中，d为天线中心点与RFID标签的直线距离，RSSI为接受信号强度(负值)，A为发射天线和RFID标签距离为1米时的RSSI值，n为环境衰减因子，上述公式经过转换可得：Where d is the linear distance between the antenna center point and the RFID tag, RSSI is the received signal strength (negative value), A is the RSSI value when the distance between the transmitting antenna and the RFID tag is 1 meter, and n is the environmental attenuation factor. The above formula is converted. Available:

RSSI＝-log(10*n*d+A)................(7)RSSI=-log(10*n*d+A).............(7)

在本实施例中，图书馆机器人沿书架边沿平行移动，故而天线与书架边沿，即到RFID标签的垂直距离保持不变，因而可以得知：In this embodiment, the library robot moves parallel along the edge of the bookshelf, so that the antenna and the edge of the bookshelf, that is, the vertical distance to the RFID tag remain unchanged, and thus can be known:

d＝L/cosα..........................................(8)d=L/cosα..........................................(8)

其中，L为天线中心到书架边沿的垂直距离，在本实施例中为5cm。α为RFID标签相对于天线中心线的偏角。带入上面的公式后可得：Where L is the vertical distance from the center of the antenna to the edge of the bookshelf, which is 5 cm in this embodiment. α is the off-angle of the RFID tag relative to the centerline of the antenna. Bring the formula above to get:

RSSI＝-log(10*n*L/cosα+A)..........(9)RSSI=-log(10*n*L/cosα+A)..........(9)

由此可知，RSSI与α成周期性变化规律，变化周期为2π。通过同时分析及实验，α与RSSI的关系有如图2的规律。It can be seen that the RSSI and α are periodically changed, and the period of change is 2π. Through simultaneous analysis and experimentation, the relationship between α and RSSI has the regularity as shown in Fig. 2.

通过图2可以得出以下结论：The following conclusions can be drawn from Figure 2:

当夹角α由大到小时，RSSI值递增；反之，有小到大时，RSSI值递减。When the angle α is from large to small, the RSSI value is incremented; otherwise, when it is small to large, the RSSI value is decremented.

根据相关资料，相位差随着距离变化规律如下：According to the relevant data, the phase difference varies with distance as follows:

其中

表示相位差，d表示距离，即阅读器天线与RFID标签的距离，λ表示电磁波的波长，δ表示其它的影响因素。相位差θ随着距离呈现周期性的变化规律，变化周期为2π。 among them

Indicates the phase difference, d represents the distance, that is, the distance between the reader antenna and the RFID tag, λ represents the wavelength of the electromagnetic wave, and δ represents other influencing factors. The phase difference θ exhibits a periodic variation with the distance, and the period of change is 2π.

同样，经公式带入后可得：Similarly, after the formula is brought in, you can get:

由此可知，

与α成周期性变化规律，变化周期为2π。通过同时分析及实验，

与α的关系有如图3的规律。 This shows that

Regularly change with α, the period of change is 2π. Through simultaneous analysis and experimentation,

The relationship with α has the law as shown in Fig. 3.

通过图3可以得出以下结论：The following conclusions can be drawn from Figure 3:

当夹角α由大到小时，相位差值递增；反之，有小到大时，相位差值递减。When the angle α is from large to small, the phase difference is increased; otherwise, when it is small to large, the phase difference is decremented.

根据以上的变化规律，通过大量实验数据的整合，基于模糊逻辑方法，可以建立一个双输入-单输出的

模糊推理***，其隶属函数如图4a、图4b、图4c所示。同时，可以得到如下的推理规则： According to the above change rules, through the integration of a large number of experimental data, based on the fuzzy logic method, a dual input-single output can be established.

The fuzzy inference system, its membership functions are shown in Figure 4a, Figure 4b, Figure 4c. At the same time, you can get the following inference rules:

If RSSI is High and

is Low Thenαis Low If RSSI is High and

Is Low Thenαis Low

If RSSI is High and

is High Thenαis Medium If RSSI is High and

Is High Thenαis Medium

If RSSI is Low and

is High Thenαis Medium If RSSI is Low and

Is High Thenαis Medium

If RSSI is Low and

is High Thenαis High If RSSI is Low and

Is High Thenαis High

基于以上的推理***和规则，在后续实验中，可以直接通过RSSI值得到α的值。由于RFID标签的位置已知，因此，可以通过RFID标签位置和α的值判断RFID读取天线的位置。在本实施例中，使用了两个天线，因而，最终图书馆机器人的位置需通过取平均值的方式获得。Based on the above inference systems and rules, in subsequent experiments, the value of α can be obtained directly from the RSSI value. Since the position of the RFID tag is known, the position of the RFID reading antenna can be judged by the RFID tag position and the value of α. In the present embodiment, two antennas are used, and thus, the position of the final library robot is obtained by averaging.

本实施例中隶属函数如下：The membership functions in this embodiment are as follows:

RSSI的隶属函数：The membership function of RSSI:

其中，μ _Low(RSSI)、μ _High(RSSI)分别表示RSSI值对于“低”(Low)和“高”(High)的隶属度。 Among them, μ _Low (RSSI) and μ _High (RSSI) respectively indicate the membership of the RSSI value for "Low" and "High".

相位差

的隶属函数： Phase difference

Membership function:

其中，

分别表示相位差

对于“低”(Low)和“高”(High)的隶属度。 among them,

Representing the phase difference

The degree of membership for "Low" and "High".

方位角ɑ的隶属函数：Membership function of azimuth ɑ:

其中，μ _Low(α)、μ _Middle(α)以及μ _High(α)分别表示方位角ɑ对于“低”(Low)、“中”(Middle)和“高”(High)的隶属度。 Among them, μ _Low (α), μ _Middle (α), and μ _High (α) represent the degrees of membership of the azimuth ɑ for “Low”, “Middle”, and “High”, respectively.

本实施例提供的基于超高频RFID技术的图书馆机器人定位导航技术，其中数据采集的过程如下：The library robot positioning and navigation technology based on UHF RFID technology provided by this embodiment, wherein the data collection process is as follows:

1、在书架边沿每隔一定距离粘贴RFID标签，并在RFID标签中录入RFID标签ID信息，同时将ID信息、标签的实际坐标以及每个RFID标签的相邻RFID标签ID信息存入后台数据库服务器。1. Attach the RFID tag at a certain distance on the edge of the bookshelf, and enter the RFID tag ID information in the RFID tag, and store the ID information, the actual coordinates of the tag, and the adjacent RFID tag ID information of each RFID tag into the background database server. .

2、图书馆机器人将沿书架边沿从一段缓慢移动向另一端，期间，图书馆机器人与书架边沿的距离保持一定距离。2. The library robot will move slowly from one segment to the other along the edge of the bookshelf, during which the library robot keeps a certain distance from the edge of the bookshelf.

图书馆机器人在移动过程中，速度保持稳定，每隔0.1s进行一次数据扫描。RFID读取设备的标号信息no，RFID标签的ID信息id，两个RFID读取设备分别采集到当前读取RFID标签的相位差

记录。 During the movement of the library robot, the speed is stable and a data scan is performed every 0.1s. The label information no of the RFID reading device, the ID information id of the RFID tag, and the two RFID reading devices respectively collect the phase difference of the currently read RFID tag

recording.

3、将采集到的数据信息保存到数据库，以备后续流程的分析和处理。3. Save the collected data information to the database for analysis and processing of subsequent processes.

采集数据后，需要对数据进行预处理，具体实现过程如下：After collecting data, you need to preprocess the data. The specific implementation process is as follows:

1、将采集到的RFID标签信息，从中筛选出id为目标RFID标签的数据。1. Collect the collected RFID tag information and filter out the data with the id as the target RFID tag.

2、由于在环境中存在许多抱歉，而RFID标签之间存在相互干扰和多径效应的硬性，造成采集的数据存在误差和噪声，从而影响后续计算的结果。设定有关RSSI的阈值threshold，当搜集的数据小于threshold时，判断为无效数据，应直接删除。2. Because there are many sorrows in the environment, and there are mutual interference and multipath effects between the RFID tags, the collected data has errors and noise, which affects the results of subsequent calculations. Set the Threshold threshold for RSSI. When the collected data is less than threshold, it is judged as invalid data and should be deleted directly.

3、对于以预处理过的数据进行平滑处理，从而取出高斯噪声的影响。3. Smoothing the preprocessed data to extract the influence of Gaussian noise.

数据进行预处理后，利用上述的推理***隶属函数，得到当前图书馆机器人的位置信息。具体实现过程如下：After the data is preprocessed, the location information of the current library robot is obtained by using the above-mentioned inference system membership function. The specific implementation process is as follows:

1、假设P _j，j＝1,2,...M代表一次扫描后RFID标签相对于图书馆机器人的可能位置，并且这些点与图书馆机器人的中心的连线之间的间隔相同，即P _j与α之间一一对应。 1. Let P _J , j = 1, 2, ... M represent the possible positions of the RFID tags relative to the library robot after one scan, and the intervals between these points and the center of the library robot are the same, ie There is a one-to-one correspondence between P _j and α.

2、当图书馆机器人移动过程中，将一段时间内的数据综合考虑，分别取RSSI和相位差的平均值作为推理***的输入值，并计算出输入变量对应的隶属度。2. When the library robot moves, consider the data in a period of time, take the average of RSSI and phase difference as the input value of the inference system, and calculate the membership degree corresponding to the input variable.

3、根据推理规则，计算出输入变量对于每条规则的隶属度。3. Calculate the membership of the input variable for each rule based on the inference rules.

4、采用加权平均法进行清晰化计算，最终得到α的清晰值。4. The weighted average method is used for the clear calculation, and finally the clear value of α is obtained.

对两组数据分别进行上述操作，分别得到Pos1＝<id ₁、α ₁>，Pos2＝<id ₂、α ₂>。在得到两组值之后，可以计算得到RFID标签相对于图书馆机器人的位置坐标，由于RFID标签的实际坐标已知，最终可以利用公式(1)-(5)得到图书馆机器人的当前位置信息<x _current,y _current>。 The above operations are performed on the two sets of data, respectively, to obtain Pos1=<id ₁ , α ₁ >, Pos2=<id ₂ , α ₂ >, respectively. After obtaining two sets of values, the position coordinates of the RFID tag relative to the library robot can be calculated. Since the actual coordinates of the RFID tag are known, the current position information of the library robot can be finally obtained by using formulas (1)-(5). x _current , y _current >.

已知图书馆机器人当前位置后，需要进行后续的导航工作，具体过程如下：After the current location of the library robot is known, subsequent navigation work is required. The specific process is as follows:

1、判断图书馆机器人当前的位置是否一定到达当前的目标位置<x _goal,y _goal>，判断标准为： 1. Determine whether the current position of the library robot must reach the current target position <x _goal , y _goal >, and the judgment criteria are:

x _goal＝x _current±ω x _goal =x _current ±ω

y _goal＝y _current±ω y _goal =y _current ±ω

即当前位置坐标与目标位置坐标的横纵差值不超过ω(cm)。That is, the horizontal and vertical differences between the current position coordinates and the target position coordinates do not exceed ω (cm).

2、若判断已达到目标位置，则将列表中保存的RFID标签ID的第一个数据删除，并更新新的目标位置为更改后的列表的第一个ID。若列表在删除后已空，则表明图书馆机器人已到达最终的目标位置，图书馆机器人停止运动；反之，通过查询数据库找到其实际坐标，更新<x _goal,y _goal>的值为相应值。 2. If it is determined that the target location has been reached, the first data of the RFID tag ID saved in the list is deleted, and the new target location is updated to be the first ID of the changed list. If the list is empty after deletion, it indicates that the library robot has reached the final target position, and the library robot stops moving; otherwise, the actual coordinates are found by querying the database, and the value of <x _goal , y _goal > is updated accordingly.

3、若判断未到达当前目标位置，则利用公式(1)-(3)计算图书馆机器人的前进方向偏角，继续移动，并重复之前的步骤。3. If it is judged that the current target position has not been reached, calculate the forward direction declination of the library robot using equations (1)-(3), continue to move, and repeat the previous steps.

本实施例的一种图书馆机器人定位导航方法流程如图5所示。A flow chart of a library robot positioning navigation method in this embodiment is shown in FIG. 5.

1、RFID标签的放置与数据绑定录入1, RFID tag placement and data binding entry

将RFID标签按规定贴在书架的边沿。尽量让RFID标签与天线中心处于同一水平线上。如果RFID天线的高度不固定，则将RFID标签贴在书架每一层的边沿。每一层的RFID标签之间间隔相同，视书架整体长度而定。由于算法精度要求，RFID标签的间隔不宜过大。例如，当RFID天线的读取功率在30dbm时，RFID标签的间隔不宜超过15cm。Attach the RFID tag to the edge of the bookshelf as specified. Try to keep the RFID tag at the same level as the center of the antenna. If the height of the RFID antenna is not fixed, the RFID tag is attached to the edge of each layer of the bookshelf. The spacing between the RFID tags on each layer is the same, depending on the overall length of the bookshelf. Due to the accuracy requirements of the algorithm, the spacing of the RFID tags should not be too large. For example, when the reading power of the RFID antenna is 30 dBm, the spacing of the RFID tags should not exceed 15 cm.

按一定规则对所有RFID标签进行编号，要求ID号包含：书架号，层数以及这一层的标号。例如，第10书架第1层第5个标签，其对应编号为“100105”。All RFID tags are numbered according to certain rules, and the ID number is required to include: the book number, the number of layers, and the label of this layer. For example, the fifth label on the first floor of the 10th bookshelf, the corresponding number is "100105".

在数据库中将RFID标签ID和对应的实际坐标一同存入。The RFID tag ID is stored in the database together with the corresponding actual coordinates.

2、利用DFS(Depth First Search)深度优先算法找到起点到终点的最短路径。利用DFS算法，根据RFID标签的相邻关系，先找到距离起点坐标最近的RFID标签Tag _start，在根据其相邻RFID标签(指处于同一书架同一层的相邻RFID标签)的坐标，不断向外拓展，直到找到距离终点坐标最近的RFID标签Tag _end，同时，将这条路径上经过的所有标签记录在列表中，并将列表中第一个RFID标签作为当前目标RFID标签，其对应的坐标作为当前目的坐标，记录在<x _goal,y _goal>中。 2. Use the DFS (Depth First Search) depth-first algorithm to find the shortest path from the start point to the end point. Using the DFS algorithm, according to the adjacent relationship of the RFID tag, first find the RFID tag Tag _start closest to the starting point coordinate, and continuously outward according to the coordinates of its adjacent RFID tag (referring to the adjacent RFID tag on the same layer of the same bookshelf) Expand until the RFID tag Tag _end closest to the end point coordinate is found. At the same time, all the tags passing through the path are recorded in the list, and the first RFID tag in the list is used as the current target RFID tag, and the corresponding coordinates are used as the coordinates. current object coordinates is recorded in <x _{_goal,} y _goal> in.

3、图书馆机器人移动同时进行RFID标签扫描3. Library robot moves simultaneously to scan RFID tags

由于RSSI值与相位差值对距离非常敏感，因而在移动过程中，图书馆机器人天线正对书架，并与书架边沿保持10cm左右的距离。同时，为保证在移动过程中RFID天线读取到足够多的数据，应将移动速度控制在一定范围内，被实施例中为0.1m/s。Since the RSSI value and the phase difference are very sensitive to the distance, during the movement, the library robot antenna is facing the bookshelf and keeps a distance of about 10 cm from the edge of the bookshelf. At the same time, in order to ensure that the RFID antenna reads enough data during the movement, the moving speed should be controlled within a certain range, which is 0.1 m/s in the embodiment.

每次读取数据后，提取出需要的信息，包括RFID标签的ID、RFID标签返回的RSSI值、RFID标签返回的相位差值以及读取的时间戳t。将读取的数据保存在后台数据库中。Each time the data is read, the required information is extracted, including the ID of the RFID tag, the RSSI value returned by the RFID tag, the phase difference returned by the RFID tag, and the timestamp t read. Save the read data in the backend database.

4、对数据进行预处理4, pre-processing the data

首先从获取的数据中，根据ID信息筛选出属于当前目标RFID标签的数据信息。再根据天线号no的不同，将新的数据分为两组。新得到的数据就是两个RFID读取天线对目标RFID标签的读取数据。对数据中的RSSI值和相位差值进行平滑处理，以去除高斯噪声带来的影响。将预处理后的数据保存在后台数据库中。First, from the acquired data, the data information belonging to the current target RFID tag is filtered according to the ID information. According to the antenna number no, the new data is divided into two groups. The newly obtained data is the reading data of the two RFID reading antennas on the target RFID tag. The RSSI value and the phase difference value in the data are smoothed to remove the influence of Gaussian noise. The preprocessed data is saved in a backend database.

5、利用模糊推理***的隶属函数获得RFID标签相对于图书馆机器人的方位角ɑ5. Using the membership function of the fuzzy inference system to obtain the azimuth of the RFID tag relative to the library robotɑ

将搜集到的数据经过平均后作为输入值带入隶属函数中，计算得到输入值对于每个元素的隶属度，从中取平均值，作为该输入量的模糊结果。将得到的RSSI值和相位差值的模糊结果输入推理规则中，计算得到隶属度最高的推理规则，从而可以得出输出量所述的模糊集合。再利用加权平均法进行去模糊化操作，得到输出变量ɑ的清晰值。After the collected data is averaged, it is taken as an input value into the membership function, and the membership degree of the input value for each element is calculated, and an average value is taken as the fuzzy result of the input quantity. The obtained fuzzy result of the RSSI value and the phase difference value are input into the inference rule, and the inference rule with the highest degree of membership is calculated, so that the fuzzy set described by the output quantity can be obtained. Then use the weighted average method to perform the defuzzification operation to obtain the clear value of the output variable ɑ.

6、利用方位角ɑ进行定位导航6, using azimuth ɑ for positioning navigation

根据两组方位角ɑ的值与目标RFID标签的实际坐标<x _goal,y _goal>的值，可以计算出当前图书馆机器人的实际坐标<x _current,y _current>。这样就实现了定位操作。根据 <x _current,y _current>与<x _goal,y _goal>之间的误差大小可以判断图书馆机器人是否到达当前目标位置。在本实施例中，设定threshold为3cm。若符合判断要求，则判断图书馆机器人已到达目标位置。将路径列表中第一个RFID标签删除，同时更新目标RFID标签为新的列表中的第一个RFID标签ID，更新<x _goal,y _goal>为其对应的坐标。若新列表中已无剩余RFID标签，则说明图书馆机器人已移动到终点位置，本次运动过程停止；否则，则利用公式(1)-(3)计算图书馆机器人新的位姿，更新后重复执行扫描步骤。 Based on the values of the two sets of azimuth angles 与 and the actual coordinates of the target RFID tag <x _goal , y _goal >, the actual coordinates of the current library robot <x _current , y _current > can be calculated. This achieves the positioning operation. According to the error between <x _current , y _current > and <x _goal , y _goal >, it can be judged whether the library robot reaches the current target position. In this embodiment, the threshold is set to be 3 cm. If the judgment requirements are met, it is judged that the library robot has reached the target position. Delete the first RFID tag in the path list, and update the target RFID tag to the first RFID tag ID in the new list, and update <x _goal , y _goal > to its corresponding coordinates. If there is no RFID tag remaining in the new list, the library robot has moved to the end position, and the motion process stops; otherwise, the new pose of the library robot is calculated using formulas (1)-(3). Repeat the scanning step.

本发明提供了一种图书馆机器人定位导航方法，具体实现该技术方案的方法和途径很多，以上所述仅是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。本实施例中未明确的各组成部分均可用现有技术加以实现。The present invention provides a library robot positioning and navigation method, and the method and the method for implementing the technical solution are numerous. The above description is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art will Several modifications and refinements can be made without departing from the principles of the invention, and such modifications and refinements are also considered to be within the scope of the invention. The components that are not clear in this embodiment can be implemented by the prior art.

Claims

一种图书馆机器人定位导航方法，其特征在于，包括如下步骤：A library robot positioning navigation method, comprising the following steps:

步骤1，建立模糊推理***，得到对应的推理规则；Step 1: Establish a fuzzy inference system to obtain a corresponding inference rule;

步骤2，将图书馆机器人预定路径中经过的RFID标签的ID信息按顺序存入列表中；Step 2, the ID information of the RFID tag passing through the predetermined path of the library robot is sequentially stored in the list;

步骤3，确定图书馆机器人下一步需到达的目标RFID标签的ID信息，并计算出图书馆机器人的前进方向，图书馆机器人开始朝下一步需到达的目标RFID标签所在位置移动；Step 3, determining the ID information of the target RFID tag that the library robot needs to reach next, and calculating the forward direction of the library robot, and the library robot starts moving toward the position of the target RFID tag to be reached next;

步骤4，图书馆机器人移动过程中，使用装置于图书馆机器人上的两个RFID读取设备分别读取放置于书架边沿的RFID标签，第一个RFID读取设备的天线记为天线1，第二个RFID读取设备的天线记为天线2，RFID读取设备读取的信号包括：RFID读取设备的标号信息no，RFID标签的ID信息id，两个RFID读取设备分别采集到当前读取RFID标签的相位差
及RFID标签的RSSI值r，以及采集到该RFID标签的时间戳t，用五元组
记录；no取值为1或2，为1时表示第一个RFID读取设备，为2时表示第二个RFID读取设备； Step 4: During the movement of the library robot, the two RFID reading devices on the library robot are respectively used to read the RFID tags placed on the edge of the bookshelf, and the antenna of the first RFID reading device is recorded as the antenna 1, the first The antennas of the two RFID reading devices are recorded as antennas 2. The signals read by the RFID reading device include: label information no of the RFID reading device, ID information id of the RFID tag, and two RFID reading devices respectively collect the current reading. Take the phase difference of the RFID tag
And the RSSI value r of the RFID tag, and the time stamp t of the RFID tag, using the quintuple
Record; no value is 1 or 2, when it is 1, it means the first RFID reading device, when it is 2, it means the second RFID reading device;

步骤5，对于五元组中的信息进行数据预处理；Step 5: Perform data preprocessing on the information in the quintuple;

步骤6，将预处理后的五元组中的信息输入步骤1得到的推理规则中，得到关于图书馆机器人的当前位置信息，并根据步骤2中的列表判断是否已经到达目标RFID标签所在位置，从而确认下一步的前进方向，以实现定位导航功能。Step 6, the information in the pre-processed quintual is input into the inference rule obtained in step 1, and the current location information about the library robot is obtained, and according to the list in step 2, it is determined whether the location of the target RFID tag has been reached. Thereby confirming the next direction of advancement to achieve the positioning navigation function.
根据权利要求1所述的方法，其特征在于，步骤1包括：The method of claim 1 wherein step 1 comprises:

步骤1-1，根据模糊逻辑方法，基于RFID信号的传播特性，建立一组双输入、单输出的模糊推理***，其中，输入变量为同一RFID读取设备对同一RFID标签读取的RFID标签RSSI值和相位差
输出变量为图书馆机器人相对于RFID标签的方向角α；并得出模糊推理***对应的隶属函数，通过隶属函数，建立起
的关系； Step 1-1, according to the fuzzy logic method, based on the propagation characteristics of the RFID signal, establish a set of dual input, single output fuzzy inference system, wherein the input variable is the RFID tag RSSI read by the same RFID reading device for the same RFID tag. Value and phase difference
The output variable is the direction angle α of the library robot relative to the RFID tag; and the membership function corresponding to the fuzzy inference system is obtained, which is established by the membership function.
Relationship;

步骤1-2，根据步骤1-1中建立的模糊推理***，得出模糊逻辑推理规则。Step 1-2, according to the fuzzy inference system established in step 1-1, the fuzzy logic inference rule is obtained.
根据权利要求2所述的方法，其特征在于，步骤2包括：The method of claim 2 wherein step 2 comprises:

步骤2-1，在书架边沿每隔一定距离粘贴RFID标签，RFID标签的ID信息包括书架号、层数以及这一层的标号，按一定规则对所有RFID标签进行编号，将RFID标签的ID信息与标签的具***置坐标信息以及标签的相邻关系进行绑定，将绑定结果存入数据库中；In step 2-1, the RFID tag is pasted at a certain distance on the edge of the bookshelf. The ID information of the RFID tag includes the shelf number, the number of layers, and the label of the layer. All RFID tags are numbered according to certain rules, and the ID information of the RFID tag is determined. Binding to the specific location coordinate information of the tag and the adjacent relationship of the tag, and storing the binding result in the database;

步骤2-2，通过查询数据库，确定距离图书馆机器人预定路径的起始位置最近的RFID标签和终止位置最近的RFID标签，并分别记为Tag _start和Tag _end； Step 2-2, by querying the database, determining the RFID tag closest to the starting position of the predetermined path of the library robot and the RFID tag closest to the end position, and respectively recording them as Tag _start and Tag _end ;

步骤2-3，通过DFS深度优先算法，寻找以Tag _start标签为起点，以Tag _end标签为终点的最短路径，并将路径上经过的所有RFID标签的ID信息按照起点到终点的顺序储存入列表中。 Step 2-3: Search for the shortest path starting from the Tag _start tag and ending with the Tag _end tag by using the DFS depth-first algorithm, and store the ID information of all RFID tags passing through the path in the order from the start point to the end point. in.
根据权利要求3所述的方法，其特征在于，步骤3包括：将步骤2所得的列表中第一个RFID标签作为图书馆机器人下一步需到达的目标RFID标签。The method according to claim 3, wherein the step 3 comprises: using the first RFID tag in the list obtained in the step 2 as the target RFID tag to be reached by the library robot.
根据权利要求4所述的方法，其特征在于，步骤5包括：The method of claim 4 wherein step 5 comprises:

步骤5-1，根据RFID标签的ID信息id的不同，将id为步骤3中所得的目标RFID标签id的Tag数据划分成新的子集，并按照时间戳t的值升序排序；Step 5-1, according to the ID information id of the RFID tag, the Tag data whose id is the target RFID tag id obtained in step 3 is divided into new subsets, and sorted in ascending order according to the value of the time stamp t;

步骤5-2，将新得到的子集按照不同的RFID读取设备标号no分成两个新的子集，每个子集内将步骤4所得的所有五元组数据Tag按时间戳t的值升序排序，将排序结果分别保存在四元组
和
中，Data1中no取值为1，Data2中no取值为2，r ₁表示通过第一个RFID读取设备采集到的当前读取RFID标签的RSSI值信息，
表示通过第一个RFID读取设备采集到的当前读取RFID 标签的相位差，r ₂表示通过第二个RFID读取设备采集到的当前读取RFID标签的RSSI值信息，
表示通过第二个RFID读取设备采集到的当前读取RFID标签的相位差。 Step 5-2, the newly obtained subset is divided into two new subsets according to different RFID reading device labels no, and all the quintuple data tags obtained in step 4 are sorted by the value of the time stamp t in each subset. Sort, save the sort results in the quad
with
The value of no in Data1 is 1, and the value of no in Data2 is 2, and r ₁ represents the RSSI value information of the currently read RFID tag collected by the first RFID reading device.
Representing the phase difference of the currently read RFID tag collected by the first RFID reading device, and r ₂ representing the RSSI value information of the currently read RFID tag collected by the second RFID reading device,
Indicates the phase difference of the currently read RFID tag collected by the second RFID reading device.
根据权利要求5所述的方法，其特征在于，步骤6包括：The method of claim 5 wherein step 6 comprises:

步骤6-1，将步骤5中所得的两个四元组Data1和Data2，依据步骤1中所得的推理规则和隶属函数，分别计算得到图书馆机器人的当前位置信息Pos ₁＝<id ₃、α ₁>，Pos ₂＝<id ₄、α ₂>，Pos ₁表示根据第一个四元组Data1计算得到的图书馆机器人的当前位置信息，其中，id ₃表示根据第一个四元组Data1计算得到的当前距离机器人最近的RFID标签的ID信息，α ₁表示根据第一个四元组Data1计算得到的当前机器人相对于该RFID标签的方位角；Pos ₂表示根据第二个四元组Data2计算得到的图书馆机器人的当前位置信息，其中，id ₄表示根据第二个四元组Data2算得到的当前距离机器人最近的RFID标签的ID信息，α ₂表示根据第二个四元组Data2计算得到的当前机器人相对于该RFID标签的方位角； Step 6-1, the two quaternions Data1 and Data2 obtained in step 5 are respectively calculated according to the inference rules and membership functions obtained in step 1, and the current position information of the library robot is calculated Pos ₁ = <id ₃ , α ₁ >, Pos ₂ =<id ₄ , α ₂ >, Pos ₁ represents the current position information of the library robot calculated according to the first quad PDF1, wherein id ₃ represents the calculation according to the first quad Data1 Obtaining the ID information of the RFID tag closest to the robot from the current distance, α ₁ represents the azimuth of the current robot relative to the RFID tag calculated according to the first quad DS1; Pos ₂ represents the calculation according to the second quad Data2 The current location information of the obtained library robot, wherein id ₄ represents the ID information of the RFID tag closest to the current distance robot calculated according to the second quaternion Data2, and α ₂ represents the calculation according to the second quaternion Data2 The azimuth of the current robot relative to the RFID tag;

步骤6-2，根据步骤6-1中所得位置信息Pos ₁和Pos ₂，依据如下公式更新图书馆机器人的方位角θ _rotation： Step 6-2, according to the position information Pos ₁ and Pos ₂ obtained in step 6-1, the azimuth angle θ _{rotation of the} library robot is updated according to the following formula:

θ _rotation＝θ _{previous->cuurent}+θ _{current->goal} θ _rotation =θ _{previous->cuurent} +θ _{current->goal}

其中，(x _previous,y _previous)为上一次定位时获得的图书馆机器人位置信息，x _previous,y _previous分别表示上一次定位时获得的图书馆机器人位置横坐标和纵坐标，(x _goal,y _goal)为图书馆机器人的目标位置，x _goal,y _goal分别表示目标位置横坐标和纵坐标，由目标RFID标签的位置确定，(x _current,y _current)为图书馆机器人当前位置，x _current,y _currentl分别表示当前位置横坐标和纵坐标，d为图书馆机器人与书架边沿间的垂直距离，δ为噪声误差，θ _{previous->current}是指当前位置与上一次定位位置之间的方向角，θ _{current->goal}是指目标位置与当前位置之间的方向角，x _id1、y _id1分别表示第一个四元组Data1计算得到的图书馆机器人的当前位置横坐标和纵坐标，x _id2、y _id2分别表示第二个四元组Data2计算得到的图书馆机器人的当前位置横坐标和纵坐标； Where (x _previous , y _previous ) is the location information of the library robot obtained during the _previous positioning, x _previous , y _previous respectively represent the horizontal and vertical coordinates of the library robot position obtained during the _previous positioning, (x _goal , y _goal) libraries for the target position of the robot, x _{_goal,} y _goal target position respectively the abscissa and ordinate, determined by the position of the target RFID tag, (x _{_current,} y _current) to the current position of the robot library, x _current, y _currentl respectively represents the horizontal and vertical coordinates of the current position, d is the vertical distance between the library robot and the edge of the bookshelf, δ is the noise error, and θ _{previous->current} refers to the direction angle between the current position and the previous positioning position. θ _{current->goal} is the direction angle between the target position and the current position, and x _id1 and y _id1 respectively represent the current position of the library robot calculated by the first quad of Data1, the abscissa and the ordinate, x _id2 , y _id2 respectively represents the current position abscissa and ordinate of the library robot calculated by the second quaternary Data2;

步骤6-3，根据步骤6-2中所得位置信息(x _current,y _current)判断图书馆机器人是否已到达步骤3所得下一步需到达的目标RFID标签所在位置，若满足(x _current,y _current)的横坐标和纵坐标的对应差值的绝对值与(x _goal,y _goal)的横坐标和纵坐标的对应差值的绝对值都小于threshold，则判断已到达目标位置，若已到达，则将步骤2中列表标记的下一步需到达的目标目标RFID标签删除，继续进行步骤6-4，否则，返回执行步骤4； Step 6-3, according to the position information (x _current , y _current ) obtained in step 6-2, it is determined whether the library robot has reached the position of the target RFID tag to be reached in the next step, if it satisfies (x _current , y _current The absolute value of the corresponding difference between the abscissa and the ordinate and the abscissa of the abscissa and the ordinate of (x _goal , y _goal ) are both smaller than threshold, and it is judged that the target position has been reached, and if it has arrived, Then delete the target target RFID tag to be reached in the next step of the list mark in step 2, and proceed to step 6-4; otherwise, return to step 4;

步骤6-4，判断步骤2所得列表中是否还存在RFID标签ID，若存在，则返回执行步骤3，否则，则判断图书馆机器人已到达预定最终位置。In step 6-4, it is determined whether the RFID tag ID still exists in the list obtained in step 2. If yes, return to step 3, otherwise, it is determined that the library robot has reached the predetermined final position.